738
L. E. EBERSON AND K. PERSON
Vol. 5
weight significantly lower than that produced by both the thiouracil control ( P > 0.99) and by 3.0 mcg. of L-thyroxine ( P > 0.95), and not significantly different from the maximal reversal produced by 6.3 mcg. of L-thyroxine. This 4’-amiiio analog of 3,3,:3’-triiodo-L-thyronine is, therefore, a t least 1.5% as active as rd-thyroxinr. Xdditional graded dose assays will be required to fix thc exact lewl of activity. The thyromimetic activity shown by two 4’-amirio analogs of thyroxine indicates that the amino group is capablr of functioning in place of the phenolic hydroxyl group in reversing thiouracil-induced goiter in the rat. This would support a mechanism of action related to electron transport, and present evidenw against a mechanisni involving simple interaction b e t w e n phenoxide ion and biological receptor. The additional requiremeiit for appropriate substitution ortho to the amino group (iodine or dimethyl) for minimal activity, when supplemented by additional examples, may help to clarify the role of the “prime ring” aiid its wbstituents in the thyroxine-like response. Acknowledgment.-This work \vas supported by a U.S.P.H.S. research grant h4223, National Institute of Arthritis and AIetaboljc Diseases, which we gratefully acknoidedge. P.S.thanks the Wellcome Trubt, London, for the an-ard of a travel scholarship.
Studies on Monoamine Oxidase Inhibitors. I. The Autoxidation of p-Phenylisopropylhydrazine as a Model Reaction for Irreversible Monoamine Oxidase Inhibition L. E. EBERSOX . ~ S D E(. PE:RSSOS Draco Laboratories, POB 242, Liind, Stceden Received Soverriher
4,1961
The cupric ion catalyzed autoxidation of 6-phenylisopropylhydrazine has been studied by preparative and kinetic methods under conditions resembling those existing in biological systems. The data obtained point to a radical mechanism for the autoxidation, initiated by transfer of one electron from the hydrazine group to a cupric ion. A mechanism for irreversible monoamine oxidase inhibition is proposed, involving transfer of one electron from the hydrazine to the enzyme and subsequent oxidation of the hydrazine radical by molecular oxygen.
July, 1962
MONOAMINE OXIDASEINHIBITORS. I
739
The intermediate radicals react with the enzyme reactive center producing irreversible changes in its structure.
The mechanism of the irreversible inhibition of monoamine oxidase (MAO) by hydrazine derivatives has been explained mainly on the basis of their nucleophilic nature.', * The electrophilic counterpart of the reactive center of the enzyme was supposed to be an "activated" carbonyl group from a peptide, amide, or ester bond, which would react with the hydrazine moiety analogously to ordinary aldehyde or ketone groups. The weaknesses of this theory are apparent. First, as hydrazone formation is a reversible r e a ~ t i o n restoration ,~ of the enzyme activity in vivo should occur considerably more rapidly than actually is the case, either by hydrolysis or exchange with highly reactive carbonyl compounds, such as pyruvic acid.4 Second, the concept of an activated peptide, amide, or ester carbonyl function capable of reacting with a hydrazine group in the way indicated above is entirely hypothetical and does not correspond to any known reaction of these groups. Third, as has been shown by D a ~ i s o n , ~ the irreversible inhibition of M A 0 by iproniazid and isopropylhydrazine requires the presence of oxygen and is decreased by glutathione, which the above theory does not predict. Davison pictured the inhibition reaction as a dehydrogenation of the hydrazine taking place a t the reactive center with the formation of reaction products which react with the enzyme, producing irreversible changes in its structure. Since we are dealing with an oxidizing enzyme and one of the most characteristic chemical properties of hydrazines, notably mono- and 1,2-disubstituted hydrazines, is their extreme sensitivity to molecular oxygen,6we considered it as a natural starting point to investigate the autoxidation of hydrazines as a possible cause of irreversible M A 0 inhibition or as a model reaction for the events occurring a t the enzymic receptor. This paper deals with a preparative and kinetic investigation of the autoxidation of P-phenylisopropylhydrazine (PIH),' one of the most potent irreversible MA0 inhibitors. The results obtained point to a radical mechanism for the inhibition, the radicals being liberated at or near (1) J. A. Carbon, W. P. Burkard. and E. A. Zeller, Helu. Chim. Acta, 41, 1883 (1958). J. Barsky, W. L. Pacha, S. Sarkar, and E. A. Zeller, J. B i d . Chem., 584, 389 (1959). (3) J. B. Conant and P. D. Bartlett, J . Am. Chem. SOC..64, 2881 (1932). (4) E. Earl Royals, "Advanced Organic Chemistry," Prentice-Hall, Inc., New York, N. Y.. 1954, p. 656. ( 5 ) A. N. Davison, Biochem. J . , 67, 316 (1957). (6) M. Lesbre, in "Trait6 de Chimie Organique," Tome XV, V. Grignard, G. Dupont, and R. Locquin, Eds., Masson et Cie, Paris, France, 1948, p. 483. ( 7 ) Pheniprazine, C a t r o n s . (2)
L. E. EBERSON AND K. PERSSW
740
Vol. 5
the reactive center of the enzyme and subsequently forming covalent, bonds with or oxidizing :t furict,ional group within it,, prwumably a thiol group. Kinetic Method and Results
The gas stoichiometry of the autoxidation of P I H was first investigated and was found to agree fairly well with the formula PIH
+ 1.501-+- products + Nf
(1)
when run in a phosphate buffer of pH 7.4 at room temperature with added Cu2+ as a catalyst. The copper salt was added to make the reaction go to completion within a reasonable time. From this formula it is evident that the kinetics of the reaction could be followed by a manometric method and a few kinetic series were also run by the Warburg technique in order to study the variation of the rate constant as a function of the oxygen pressure above the solution. However, as the Warburg method is rather time-consuming and t>hc rate constants obtained were poorly reproducible, we elaborat'ed a rapid and reliable method for following the kinetics of the autoxidation. The hydrazine exists partly as the hydrazinium ion, RNHKH3" at pH 7.4 and thus the reaction can be followed by measuring the rat'e of proton liberat,ion.* This can be done conveniently by the pH-stat method, and preliminary experiments indicated that the end value of added hydroxide corresponded exactly to the yalue calculated for t'he liberation of one proton/molecule of oxidized PIH chloride. The appearance of a typical run is shown in Figure I . The pseudo firstorder rate constant, keXp= ( l / t j In [a/(n- 2 ) ] was determined graphically from a plot of log (a - ). against 1. hssuming that the free base is the kinetically active species it can be derived easily that k,,, is related to the rate constant k for the aut'oxidation of the free base through the equation (2) : log k = log
kexp
+ pK,'
-
pH
(2)
where pK,' is the apparent dissociation constant of the equilibrium (3). RNHNHa+
+ HzO Jr RNHNHz + H30+
(3)
(8) It is assumed t h a t the rate of complex formation between cupric ion and the hydrazine compound is extremely rapid compared t o the rate of autoxidation in the p H range studied. This could be shown semiquantitatively i n the following way: A small volume of cupric sulfate solution was added as fast a8 possible t o a well stirred, nitrogen flushed solution of P I H maintained a t a fixed p H by the pH-stat. Even when the pH-stat was operated a t a high speed
(0.3 ml. of 0.1 M sodium hydroxide/min.) the rate of proton liberation due to the reaction
CU'+
+ R?JHNHa+ + RNHNHz
was too rapid to be followed. With [ P I H ] = 10-1 M and [Cu2+] = 7.5 X lo-,' M this cnt'reaponds to a rate constant for complex formation >400 I, mole-' rnin.-'.
MONOAMINE OXIDASE INHIBITORS. I
July, 1962
74 1
20 -
15-
21 1 0 .-
E
5-
-
0.5
-
rnl 0.1025M NaOH
Fig. 1.-Appearance of a typical primary curve obtained by the pH-stat method. The straight line D represents the end value of added hydroxide, calculated on the basis that one proton/molecule of PIH chloride is liberated; [PIHt,t,l] = 1.71 X 10-3 M, [Cu2+] = 1.54 x 10-5 .If. For details, cf. Experimental.
The pH-stat method was used for studying the kinetics of the autoxidation of P I H in aqueous solution in the presence of catalytic amounts of cupric ion and was found to give reproducible first-order rate constant k,,, provided these conditions were fulfilled: (1) The temperature was controlled within =t0.loby the use of a thermostated reaction vessel. (2) The reaction was run in 0.15 M sodium perchlorate in order to maintain a constant ionic strength during the run. (3) The oxygen concentration of the solution was kept constant by blowing in oxygen slowly through a fritted glass disk. (4) The pH was kept a t 6.50 instead of the physiological value 7.4, as the accuracy of the rate constant calculations was improved if the experimental curve in Fig. 1 was run over a large interval (a in Fig. 1).
L. E. EBERSON ASD K. PERSSOX
743
Vol. 5
( 5 ) The autoxidation is specifically catalyzed by cupric ion. There occurs a slow autoxidation ( k - 0.02 miii.-l) in the absence of added cupric ion, which probably is due to trace impurities of cupric ion in the water and chemicals used for preparing the solutions. To obtain reasonable rates of autoxidation catalytic amounts of cupric ion must be added and after some preliminary experiments the cupric ion concentration \vas fixed a t 7.70 x JI and the total concentration of -11. In all runs the ratio [Cu2r]'[Hydrazinetotai] I'IH at 1.76 X was kept constant except in the case when the dependence of [Cu2+] \vas studied. The reproducibility of the method is shown in Table I, which gives the values of kexp and k from six runs at 37.0' and pH 6.d0 in 0.15 .I1 aqueous sodium perchlorate. Another demonstration of the accuracy is shown in Table I1 and Fig. 2, which give the results of measurements carried out a t different pH values. The equation of the straight line in Fig. 2 was calculated by means of the least squares method and found to be T.4BLE 1 RATECONST.~NTS FOR THE AUTOXIDATION O F PIH AT 37.0" AND PH 6.50 I N 0.15 -21 NACLO,~;[PIHt,t,,] = 1.76 X 32, [Cr2+] = 7.70 X 10-6 31
12 in m h - 1 calcd. from formula (2)'
kCXL>102 in n ~ i n .-:
SunibPr of r u n
5.50 0.24 6.1!1 0.27 6.11 0.27 5.66 0.25 5 6.46 0.28 6 5.55 0.24 Mean value 0.26 Standard deviation 1 0 . 0 2 The value 7.14 1~3,sused for pKa' (see Table TT). 1 2 3 4
a
log k,,,
=
-7.63
+ 0.99pH
which was in good agreement with that calculated from formula (2) using the mean value of k in Tables I and 11, 0.26 min.-', and pK,' = 7.14 log k,,,
=
-7.72
+ pH
Thus the free base must be the kinetically active species in the reaction. Table I11 shows the variation of the rate constant as a function of the cupric ion concentration. Within the limits of error the rate constant is directly proportional to [Cu2+]in t,he concentration interval
MONOAMINE OXIDASE INHIBITORS. I
July, 1962
743
TABLE I1 RATE CONSTANTS FOR THE AUTOXIDATION OF P I H AS A FUNCTION OF PH AT 37.0" IN 0.15 M NACLO~;[ P I H ~ , ~ , I=] 1.76 X 10- M , [Cr*+]= 7.70 X lopeAI lop k
PH
5.50 5.75 6.00 6.25 6.50 6.75 7.00 7.25 7.50
a
+1
lo2
calcd. from
i n min.-I
foriniila 12)"
kexp
k in min. - 1
0.66 0.98 1.98 3.23 5.9 11.2 21 34 52
0.46 0.29 0.38 0.24 0.44 0.27 0.40 0.25 0.41 0.26 0.44 0.27 0.46 0.29 0.42 0.26 0.36 0.23 Mean value 0.26 Standard deviation 1 0 . 0 2 The value 7.14 was used for pK,' (see Table V).
1.51'5.
i
lol? a 2 -* 07
0 0.5 -
0.0-
L
5.00
6.00
PH
-
7.00
E
IO
Fig. 2.-Representation of log k,,, against pH (see Table 11). The equat,ion of the straight line is log k,,, = -7.63 O.99pH.
+
I,. E. EBERSON AND K. PERSSON
744
VOl. 5
studied. The catalytic potency of some other heavy metal ions was also investigated, but none of them (Fe3+, Mn?', CO?+,Ni2+) had i~iiyappreciable influence on the autoxidation rate even in consider:hly higher concentrations.g TABLEI11
RATECONSTANTS FOR
37.0' ICu''1
X lo6 AI
PIH AS A FUKCTIOX O F [CTJ''] PH 6.50 IN 0.15 M NACLO~ k e x p x 10'
THE -4GTOXlDATION O F AND
[PIHtOtail X 103 M
1.54 3.08 4.62 6.15 7.70 9.23 10.8 12.3 13.9 15.4
1.73 1.83 1.80 1.76 1.76 I .69 1.71 1.72 I .79 1.71
in min. -1
AT
12 in m h - 1
1.30 "67 :I . 7
0.057 0.12 0.16 0.23 0.26 0.28 0.36 0.42 0.49 0.56
R ,?
5.9 6.5 8.3 9.5 11.2 12.9
If substances which can form very stable complexes with cupric ion were added the autoxidation was strongly inhibited, as is shown in Table IV, which gives the rate constants of the cupric ion-catalyzed reaction in the presence of propyl gallate. TABLEIV RATECONSTANTS FOR THE -4IJTOXIDATION O F PIH I N THE PRESENCE O F GALLATE AT 37.0" AND PH 6.50 I N 0.15 M NACLO~;[ P I H t o t a l ] = 1.76 X [Cv2+] = 7.70 X 10- 6X [Propyl gallrttel X
106 M
k,,, 10' in min.-l
0 1.75 3.50 5.25
5 9 4.5 3.6 1.97
PROPYL
M,
k in inin.-'
0.26 0 20 0.16 0.086
Table V gives the rate constants k,,, and IC and pK,' as a function of the temperature. From these values the energy of act'ivation was calculated to be 17 kcal./mole by the method of least squares. Addition of chloride and bromide had a very strong influence on the rate of autoxidation, presumably by altering the oxidation-reduction potential of the Cu2+/Cu+ system.1° Thus, in 0.15 M aqueous sodium chloride or bromide the reaction rate was too high to be meas(9) Cf.L. F. Audrieth and P. M. Mohr, Ind. and En& Chem., 18,1774 (1951). (10) W. M. Latimer, "Oxidation Potentials,'' Prentice-Hall, Ino., Englewood Cliffs, N. J., 1952, D. 186.
MONOAMINE OXIDASEINHIBITORS.I
July, 1962
745
TABLEV RATECONSTANTS FOR THE AUTOXIDATION OF PIH AS A FUNCTION OF TEMPERATURE AT PH 6.50 IN 0.15 M NACLO,; [PIHt,t,l] = 1.76 X M,[CU*+] = 7.70 x 10-6 M T.OK.
(1/T) 10'
293.16 298.16 303.16 308.16 310.16 313.16
3.411 3.354 3.299 3.245 3.224 3.193
PK.'
7.31 7.26 7.21 7.16 7.14 7.08
hex* 102 in min. -1
log k
k in niin. -.
0.76 1.39 2.5 4.8 5.9 7.9
0.69-2 0.90-2 0.11-1 0.34-1 0.41-1 0.48-1
0.049 0,080 0.13 0.22 0.26 0.30
TABLEVI RATECONSTANTS FOR THE AUTOXIDATION OF PIH AS A FUNCTION OF CHLORIDE M , [Cua+] = IONCONCENTRATION ~ ~ 3 7 . AND 0 ' ~ H 6 . 5 0 [PIHtotal] ; = 1.76 X 7.70 X 10-8 M; NO SODIUM PERCHLORATE ADDED kexp
[Cl-I, M
a
10'
in min.-1
0.00176 0.00226 0.00276 0.00325 0.00426 0.00676 At these salt concentrations a pK,'
9.9 13.2 17.2 17.8 19.1 24.6 of 7.00 was used.
k in min. - 1
a
0.31 0.42 0.54 0.56 0.60 0.78
ured by the pH-stat method, as the rate of hydroxide addition could not be increased enough. The rate constants at low chloride concentrations are shown in Table VI. The rate dependence of the oxygen pressure above the solution was determined by the Warburg method and the results are shown in Fig. 3, where the rate constants k for autoxidations in air and oxygen are represented against [Cu2+]. The straight lines in Fig. 3 are identical within the limits of the experimental error, indicating that the rate is independent of the oxygen concentration. This could also be shown by the pH-stat method, if air was blown into the solution instead of oxygen, in which case the rate constant (mean value from three runs) was 0.23 min.-1. This value agrees fairly well with that obtained with oxygen, 0.26 min.-l. Unfortunately the rate constants determined by the Warburg method are not comparable with those determined by the pH-stat method, as in the manometric experiments the reaction had to be run in a phosphate buffer. Finally the rate of autoxidation and pK,' of a number of other siibstituted hydrazines have been determined (Table VII). These
L. E. EBERSON AND I(.PERSON
746
1
Y
0.0
2 .o [CU2+].106
4.0
y
Vol. 5
6 .O
8.0
Fig. 3.-Representation of k against [Cuz-] for autoxidations in air and oxygen (see Table VII): 0 in air, A in oxygen; medium was 0.067.Mphosphate .ll. buffer of pH 6 50: temp. 37.0": [PIH] = 1.65 X
TABLE VI1 RATECONSTANTSFOR THE .%VTOXIDAkT1OK O F St-BSTITUTED HyDR4ZINES RXHSH? AT 37.0" A I ~ DPH 6.50 IN 0.15Jf SAC LO^; [RSHXH, = 1.76 X df, [Ccz-] = 7.70 X M R
Isopropyl p-Cyclohexylisopropyl Phenyl a t pH 4.50 Benzyl Phenylethyl 6-Phenylisopropy 1 p-( 2-Methylphenpl)-isopropyl p-( 4-Methoxyphenyl)-isopropyl p-( 3,4-Met~hylenedioxyphenyl)-isopropyl p-( 3,4,5-Trimethoxyphenyl)-isopropyl
Dh'rl'
7.48 7.78 5 14 6 81 7.12 7.14 7 08
kexI, l(J! in rnin. - 1
0.93 2.0 15.1
i.09
A8 9.8 5.0 5.3 3.2 3.5
7.03
6.2
7.17
Iz in niin. -1
0.09 0.49 0.66 0.14 0.41 0.26 0.20 0.24 0.14 0.21
include the therapeutically employed isopropyl-," benzyl-, I? and phenylethylhydra~ine,~~ and it is obvious that these compounds undergo autoxidation at rates comparable with that of PIH. A semiquantitative study on iproniazid revealed that it was also autoxidized under the conditions employed in this investigation and that the (11) The hydrazine part of N'-isonicotinoyl-N*-isopropylhydrazme, iproniazid. (12) The hydrazine part of N'-henzyl-N*-(5-methyl-3-1soxazolylcarhon~ 1)-hydraelne, carhoxazide. (13) Phenelzine.
130-
July, 1962
MONOAMINE OXIDASE INHIBITORS. I
747
reaction displayed the same characteristics, ie., it was strongly dependent on cupric and chloride ion concentration. l 4 Hydrazine itself was found to undergo rapid autoxidation under the conditions employed, but the kinetics deviated considerably from first-order behavior. Preparative Results.-The reaction products from the autoxidation of a large sample of P I H run in a phosphate buffer of pH 7.4 at room temperature in the presence of catalytic amounts of cupric ion mere isolated and crudely separated by fractional distillation. The fractions were analyzed by the infrared and gas chromatographic technique and found to consist of n-propylbenzene (I), propenylbenzene (11), phenylacetone (111), and 1-phenyl-2-propanol (pphenyl-2-propanol) (IV). The distillation residue was a black tar, from which no defined compound could be isolated. The amounts of I, 11, 111, and I T corresponded approximately to yields of 10, 10, 15, and 5% respectively, whereas the residue accounted for 30-40%. In order to obtain more evidence for the radical nature of the autoxidation, methyl methacrylate in large excess was treated with P I H and cupric ion in aqueous ethanolic solution in the presence of oxygen. The methyl methacrylate quantitatively polymerized to a solid polymer, whereas no polymerization occurred even after storage for several weeks of a solution containing the same components except PIH. In a semiquantitative kinetic experiment it was shown that peroxidic material, presumably hydrogen peroxide, accumulated during the autoxidation process. Discussion of Results.-It is well known that many autoxidations proceed according to a chain mechanism involving free radicals, 16-17 and therefore it is not surprising that the autoxidation of P I H follows the pattern of a radical reaction. The data presented in this paper do not allow a complete mechanism to be elucidated but it probably involves these steps: (1) A complex between P I H and Cu2+is formed, which decomposes according to
(14) B. Koechlin and V. Iliev, Ann. A;. Y . Acad. Sei., 80, 868 (1959). (15) C. Walling, “Free Radicals in Solution,” John Wiley & Sons, Inc., New York, N. Y . . 1957, p. 397. (16) W. A. Waters, “The Chemistry of Free Radicals,” Oxford University Press, London, 1948, p. 232. (17) W. A. Waters, in “Progress in Organic Chemistry,” P a r t 5, J. W. Cook a n d W. Car. ruthers, Eds., Butterworth & Co. Ltd., London, 1961, p. 17.
L. E. EBERSON AND K. PERSSON
748
Vol. 5
( 2 ) The radical T T is oxidized by oxygeri n-ith the forniatioii of hydrogen peroxide and the radicals VI and VII. (‘6H5CH.
CaHaCH
1.
\
+ 0- --+
/
c=\--
S1-I.
+
\
(’Hp-Y-yNH
/
+ H?O-
( 5I
CH,
CH3
\I
T IT
( 3 ) VI then gives rise to phenylacetoiie (111)and a hydrazine radical by rapid hydrolysis whereas T’II may capture an electron from a cuprous ion and become further oxidized by oxygen according to CsH,CH,
1-11
+ C U T ----f
\
/
+ CuZT+ CP,H~CIII~--CHCH~+ 0 2
CH--S=?;H
CH3
YIII
TJ+ HOO.
((3)
The radical VI11 may react in several ways, dependent on the coiicentration of the hydrazine compound. At relatively high concentrations, as in the preparative experiments, coupling or disproportionation will domiuate, while iii Iht. coticaentrations crnployed in the kinetic experiments or iinder biulogi(d cmiditioiis rcwtioii with solvent molecules or other substaureh presrnt is ~iiorelikely to occur. This reaction scheme is snpportcd by the folloning facts: Cupric. ion is known to form complexes with hydrazinc,. hiit unfortunately 110 complex constants havc hcrn rc.poi,tcd for such romplrxes. 1s,19 However, :I comparison hetwcen the ammonia and hydrazine complexes of Cd2+, Zn2+ and Si2’i1~0wl-qthat the qtnbility constanti K1 and K a for the firit two c~oinpl~.xr,i at f’ approximately equal.1x It seems reasonable to assunif’ that log ICl mid log K z f o i t h i ~complexri of t hc between cupric3 ion and hydraiiic, oi rnoiio:dkylhydi :ixiiiri kame order of magnitude as thow of the :Lmnioiiia complexes, 4.:3 and 3 7. Using the Talues log Iil = log Kz = 4 it can be calculated that complexes between cupric ion m d PI“ should dominate owl free cupric ion in the concentration iiiterval used in this investigation. The proposal that the decomposition of the complex according t o reaction (4)is the rate determining step is consistent with the kinetic results, i.e , the rate constarit i T approxim:itelp proportional to the concent ration of cupric ion and indepelident of the oxygen cuncentrn(18) J Bjerrum G Schwarzenbach, and L G SI116n. ”Stabilit) Constants of Metal-ion Complexes, P a r t I1 (Inorgsnic Ligands and Solnbilltv PIodticts) ,” Chem Soc London, Burhngton House, 1958 (19) L. F. hudrieth a n d B A. Ogg, “The C11c rnlbtrl of Hsdrazlnc ” Jolin Wlley fi < o m , Inc S e i \ T o r k , N Y , 1951, p. 190.
July, 1962
MONOAMINE OXIDASE INHIBITORS. I
749
tion. Furthermore, the autoxidation is inhibited by reagents which form strong complexes with cupric ion and accelerated by anions which stabilize the cuprous state, such as chloride and bromide ion (Table VI).'O An analogous initial step has been postulated in the autoxidation of dialkylhydroxylamines in alkaline solution in the presence of cupric ion.20*21Once formed, the radical V is very reactive toward oxygen and undergoes the reaction steps outlined in the equations ( 5 ) and (B), which in a rough manner account for the products isolated and identified. The radical nature of the autoxidation is also shown by its capability to initiate the polymerization of a large excess of methyl methacrylate, which is rather insensitive to ionic polymerization mechanisms. Radical mechanisms have been proposed for a number of oxidative processes in which hydrazine derivatives or related compounds take part, e.g., in the autoxidation of hydrazones,*2 in the oxidation of hydrazine by ferric i0n,2~,2~ and in the oxidation of phenylhydrazine by silver oxide.26 Likewise, the oxidation of hydrazones in the presence of cupric ion has been found to initiate polymerization of several ethylenically unsaturated compounds.26 As it can be shown that P I H is oxidized in the presence of oxygen also in biological system^,^' e.g., blood, the formation of radicals during this process has some important biochemical implications. Radicals are highly reactive and react rather unspecifically in a number of ways with biological structures, giving rise to irreversible changes by virtue of their ability to act as oxidizing agents, to form covalent bonds or to initiate chain reactions. There are several indications that this is what happens when hydrazine derivatives interact with biological systems. When phenylhydrazine is treated with oxygen in the presence of oxyhemoglobin benzene and nitrogen are formed, while the oxyhemoglobin is irreversibly changed.28 These findings point to an oxidation-reduction system between the oxyhemoglobin and phenylhydrazine, with hydrogen transfer from the latter to molecular oxygen. Hemin has been found to oxidize isoniazid to 1,2-diisonicotinoylhydrazineand isonicotinic acid in air at (20) D. H. Johnson, M. A. Thorold Rogers, and G . Trappe, J . Chem. Soc., 1093 (1956). (21) Ref. 14, p. 416. (22) K. H. Pausacker, J . Chem. Soc., 3479 (1950). (23) W. C. E. Higginson and P. Wright, J . Chem. Soc., 1551 (1955), and preceding papers of the series. (24) J. W. Cahn and R. E. Powell, J . A m . Chem. Soc., 76, 2568 (1954). (25) R. L. Hardie and R. H. Thomson, J . Chem. Soc., 2512 (1957). (26) U. S. Patent 2,686,775; c/. C. A . , 48, 142941 (1954). (27) L. E. Eberson and K. Persson, to be published. (28) G. H. Beaven and J. C. White, Nature, 1'78, 389 (1954).
750
L. E. EBERSON AND K. PERSEON
VOl. 5
pH 7.5, presumably vza radical intermediate^.^^ Phenylhydrazine, C14-labelledin the ring, is excreted only to (30% after 10 days when administered orally to rabbits,30 suggesting that the rest has been irreversibly attached to various sites in the body. The irreversible inhibition of M A 0 by hydrazine derivatives is analogous to their copper-catalyzed autoxidation in home important ways. First, in view of their similarity to the natural substrates of MAO, the hydrazine-type MAO-inhibitors have an affinity for the reactive center of the enzyme, which may be considered analogous to their ability to form complexes with cupric ion. -11.40 has an oxidation-reduction potential which lies between -0.05 arid $0.195 v . , ~ ~ whereas the system Cu2+/Cu+has the potential +0.133 v. By transfer of one electron from the hydrazine nitrogen to -1IA0 analogous to the transfer of an electron within the cupric ion-hydrazine complex, the radical TT is formed. By reaction of T- with molecular oxygen a reaction sequence similar to ( 5 ) and (6) involving radical intermediates is possible, and as the radicals are liberated at the reactive center of the enzyme, they dominantly react with structures within it. Since MA0 has been shown to be a thiol enzyme, it beems likely that this group is involved in the inhibition reaction, as thiol groups are easily attacked by radicals. This mechanism for irreversible M A 0 inhibition is in agreement with Davison’s finding^,^ but it cannot be excluded that the autoxidation itself is responsible for the inhibition. I n this case the radicals would be liberated in the solvent bulk outside the enzyme and react with it in an unspecific manner. However, we favor the first mechanism, as this offers an explanation for the stereospecificity found for example in the inhibition of MA0 with u- and L-forms of PIH.32 It may be of interest to mention that the amine oxidase of peaseedlings contains cupric ion which appears to be important for the enzymic activity.3 3
Experimental Materials.-The hydrazines were kindly supplied by Dr. John Biel, Lakeside Laboratories, Milwaukee, Wis., and used as hydrochlorides. Ail other chemicals were of analytical grade and the water used was distilled and passed through a demineralizing apparatus. (29) (30) (31) Jucker, (32) (33)
A. Albert and C. W. Rees. Biochem. J . , 61, 128 (1955). W.&I. McIsaac, D. V. Parke, a n d R. T. Williams, Biochem. J., TO, 688 (1958). A. Pletscher, K. F. Gey, a n d P. Zeller in “Progress in Drug Research,” P a r t 2 , E. Birkhauser Verlag. Base1 und Stuttgart, 1960, p. 427. J. Bernstein, K. A. Losee, C. Smith, and B. Rubin, J. Am. Chem. Soc., 81, 4433 (1959). P. J. G. Mann, Biochem. J., 79, 623 (1961).
July, 1962
MONOAMINE OXIDASE INHIBITORS. I
751
Determination of the Gas Stoichiometry.a4-The gas stoichiometry was determined by conventional gas-buret technique, and a ratio between absorbed oxygen and evolved nitrogen of 1.49 was obtained. Kinetics.-The pH-stat was the type SBR2c Titrigraph combined with the type SBUla syringe buret, the type TTTlc Titrator and the type TTA2 titration assembly from Radiometer, Copenhagen, Denmark. The reaction vessel was equipped with a jacket, through which thermostated water from an ultrathermostat was circulated a t a speed of about 10 l./min. The temperature constancy was better than f0.1'. The substituted hydrazine hydrochloride (4.58 x 10-6 mole) was dissolved in 26.0 ml. of 0.15 M aqueous sodium perchlorate in the reaction vessel and the mixture was allowed t o reach thermal equilibrium (point A in Fig. 1). Then the pH-stat was started and sodium hydroxide (0.1025 M ) was fed into the solution until the desired pH was obtained (point B in Fig. 1). At B the oxygen inlet was introduced into the reaction vessel and a t C the reaction was started by adding the appropriate amount of 2.00 mM cupric sulfate solution by means of a constriction micropipet. The end value of added sodium hydroxide (line D in Fig. 1) always agreed very well with that calculated for the liberation of one proton/molecule of substituted hydrazine chloride. The rate constant, k,,,, was calculated by means of the usual formula k,,, = ( l / t ) In [ a / ( a - z)] where a and ( a - 2) had the meaning which is indicated in Fig. 1. All rate constants given in Tables 11-VI11 are the mean values obtained from a t least two runs. As is seen from Tables I and I1 the relative standard error is about 2 7 % . The Warburg experiments were made under the same conditions as the pH-stat experiments, except that the medium was a phosphate buffer. The rate constants were calculated by the formula k,,, X t = In [V,/( V , - Vt)] and transformed into the rate constants k for the autoxidation of the free base. The apparent dissociation constants pK,' for the equilibrium (3) were determined from the half-neutralization points of the titration curves, and each pK,' was the mean value from four determinations. Autoxidation of P1H.-PIH chloride, 18.7 g. 10.1 mole), was dissolved in 600 ml. of 0.6 M aqueous phosphate buffer of pH 7.4. The solution was poured into a 5-1. round-bottomed flask, which previously had been flushed with oxygen. A solution of 0.1 g. of cupric sulfate in a small volume of water was added, the flask was stoppered and stirred by means of a magnetic stirrer for 24 hr. a t room temperature. The reaction mixture was acidified with concentrated hydrochloric acid and extracted with three portions of ether. The ether extracts were combined and washed with sodium carbonate solution and water and finally dried with anhydrous calcium sulfate. The sodium carbonate solution was acidified but no acidic products were obtained. The ether solution was distilled through a semi-micro distillation column of about 8 theoretical plates, and these fractions were collected: I, 1.1 g., b.p. 506O0(15mm.); 11, 1.4g., b.p.60-98"(15mm.); III,2.3g., b.p.98-99"(15mm.); IV, 0.6 g., b.p. 75-100" (0.1 mm.). The distillation residue was a black tar (5.1 g.) from which no defined compound could be isolated. By means of infrared spectra and gas chromatogram^^^ it was found that I was mainly n-propyl(34) The authors are indebted to Dr. E. Bladh, Department of Analytical Chemistry, University of Lund, for this investigation. (35) The authors wish t o express their thanks to Dr. K. Rosengren, Thermochemical Institute, University of Lund, for his kind help with the gas chromatographic investigation.
752
Pin, GOULD,LOBECK, ROTH,AND FELDKAMP
VOl. 5
benzene, I1 a mixture of n-propylbenzene and propenylbenzene, I11 phenylacetone and IV 1-phenyl-2-propanol. Phenylacetone was also identified by conversion into its dinitrophenylhydrazone and 1-phenyl-2-propanol by conversion into its 3,5-dinitrobeneoate. Minor an~ountsof other compounds were also present but no attempts t o identify them were made. The acidic solution from the ether extraction was made alkaline and continuously extracted with ether for 12 hr. The ether solution was washed with a small volume of water. dried nith anhydrous potassium carbonate and distilled. PIH, 1.1 g., b.p. 70-75" (0.02 mm.) n-as recovered, identified by conversion into the hydrochloride. Polymerization of Methyl Methacrylate.-Methyl methacrylate (8.0 g., 0.08 mole), P I H (0.25 g., 0.0013mole) and cupric chloride (0.2 mg., 1.2 X mole) were dissolved in a mixture of water (40 g . ) and ethanol (55 g.). The solution was shaken overnight after which time a solid polymer had precipitated, which n-as collected, washed with water and dried. It weighed 8.0 g. (10070yield). A control solution, which had the same composition except that P I H was omitted, had not polymerized after standing for 3 weeks.
Pyrrolidines. VII. 3-Hydroxy-1-PyrrolidinecarboxylicAcid Esters YAO-HUAWu, WILLIAMA. GOVLD,ITALTERG. LOBECK,JR., HERBERT R. ROTH,AND R. F.FELDK.4MP Mead Johnson Research Center, Evansville, Indiana 1Zeceived Jnnuaiy 25, 1862 N-Alkoxycarbonylamino acid esters were condensed with ethyl esters of a,?unsaturated carboxylic acids in the presence of sodium hydride and dry benzene to form alkyl ethyl 4-oxo-1,3-pyrrolidinedicarboxylates,which were hydrolyzed and decarboxylated to the corresponding 3-oxo-1-pyrrolidinecarboxylic acid esters. The carbonyl groups a t the 3-position of these esters were reacted selectively with appropriate Grignard reagents to yield various .?-substituted R-hyclroxy-1-pyrrolidinecarboxylic acid esters, which were submitted for scrcening as hypnotic agents.
Vor it drug produciiig central wtivit,y an adequate lipid-water solubility ratio is important in order t'o enable it t o reach the site of its a c t i o ~ i . ' ~With ~ this hypothesis in mind we prepared a. series of compounds of st'ructure (1) for screening as hypnotic agents, ( 1 ) T. Sollmann, "A Manual of Pharmacology and I t s Application t o Therapeutics and Toxicology," W. B. Saunders Company, Philadelphia, Pa., 1957, p. 826. (2) R. B. Barlow, "Introduction to Chemical Pliarmacology," John TViley & Sons, Inc., N e w Tork, N.Y., 1955, p. 29.
